Counter Electrode Impact on Quantum Dot Solar Cell Efficiencies

Sep 22, 2016 - The counter electrode (CE), despite being as relevant as the photoanode in a quantum dot solar cell (QDSC), has hardly received the sci...
2 downloads 10 Views 12MB Size
Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

The Counter Electrode Impact on Quantum Dot Solar Cell Efficiencies P. Naresh Kumar, Ankita Kolay, Sarode Krishna Kumar, Prabir K Patra, Ashish N Aphale, Avanish Kumar Srivastava, and Melepurath Deepa ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b08921 • Publication Date (Web): 22 Sep 2016 Downloaded from http://pubs.acs.org on September 22, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

The Counter Electrode Impact on Quantum Dot Solar Cell Efficiencies P. Naresh Kumara, Ankita Kolaya, S. Krishna Kumara, Prabir Patrab,c, Ashish Aphaleb, Avanish Kumar Srivastavad, Melepurath Deepaa,* a

Department of Chemistry, Indian Institute of Technology Hyderabad, Kandi-502285 Sangareddy, Telangana (India)

b

c

Department of Biomedical Engineering, Department of Mechanical Engineering, Bridgeport, CT 06604

d

b,c

University of Bridgeport,

CSIR-National Physical Laboratory, Dr. K. S. Krishnan road, New Delhi-110012 (India)

KEYWORDS Quantum dots, liquid junction, solar cells, counter electrode, efficiency

ABSTRACT: The counter electrode (CE), despite being as relevant as the photoanode in a quantum dot solar cell (QDSC), has hardly received the scientific attention it deserves. In this study, nine CEs: single walled carbon nanotubes (SWCNTs), tungsten oxide (WO3), poly(3,4-ethylenedioxythiophene) (PEDOT), copper sulfide (Cu2S), candle soot, functionalized multiwalled carbon nanotubes ( F- MWCNTs), reduced tungsten oxide (WO3-x), carbon fabric (C-Fabric), and C-Fabric/WO3-x were prepared by using low cost components and facile procedures. QDSCs were fabricated with a TiO2/CdS film which served as a common photoanode for all CEs. The power conversion efficiencies (PCEs) were: 2.02, 2.1, 2.79, 2.88, 2.95, 3.78, 3.66, 3.96, and 4.6 % respectively, and the incident photon to current conversion efficiency response was also found to complement the PCE response. Among all CEs employed here, the C-Fabric/WO3-x outperforms all the other CEs, for the synergy between C-Fabric and WO3-x comes to the fore during cell operation. The (i) low sheet resistance of C-Fabric, and its’ high surface area due to the mesh like morphology, enables high WO3-x loading during electrodeposition, and (ii) the good electrocatalytic activity of WO3-x, the very low overpotential and its high electrical conductivity that facilitate electron transfer to the electrolyte, are responsible for the superior PCE. WO3 based electrodes have not been used till date in QDSCs; the ease of fabrication of WO3 films, and their good chemical stability and scalability also favor their application to QDSCs. Futuristic possibilities for other novel composite CEs are also discussed. We anticipate this study to be useful for a well-rounded development of high performance QDSCs.

Introduction Recent advances in quantum dot solar cells (QDSCs) have propelled them to the forefront of photovoltaic research, for power conversion efficiencies (PCEs) in excess of 11% have been achieved by Zhong’s group.1 The use of passivation layers of SiO2 and ZnS in the photoanode to minimize excited electron recombination has emerged to be a powerful strategy for reducing photocurrent losses and boosting PCEs in these reports.2 The current impressive forays made in designing photoanodes for improved efficiencies, provides the stimulus to re-visit the other half of the cell, namely, the counter electrode (CE) and how it impacts PCEs. The electrolyte typically employed in a QDSC is a polysulfide ion based solution or gel. The prerequisites that a potential CE needs to satisfy to be

employable in a QDSC are listed here. (i) The CE should be highly electrocatalytic and catalyze the reduction of the polysulfide species to the polysulfide anion. (ii) It should not undergo photodecomposition. (iii) It should be chemically compatible with the polysulfide electrolyte, and should not dissolve in the electrolyte. (iv) It should be a good electron conductor conducive for fast electron transfer to the electrolyte. (v) It should have a work function in the range of 4.3-5.3 eV. (vi) Ideally, it should be inexpensive, non-toxic and earth-abundant as well. (vii) It should be optically transparent, if the QDSC is to be illuminated from the front side (optional). A variety of materials which fulfil some or most of the above described requirements have been employed in QDSCs with reasonable success in the past. Metal sulfides such as NiS, CoS, CuS, Cu2S, and PbS, conducting polymers such as

1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

poly(3,4-ethylenedixoythiophene) (PEDOT) and poly(pyrrole) (PPy), metals such as Pt, Au and Ag, and graphitic nanostructures such as multiwalled carbon nanotubes (MWCNTs), reduced graphene oxide (rGO), single walled carbon nanotubes (SWCNTs) have been used as CEs in QDSCs.3-7 Composites of Cu2S/RGO, Au/MWCNTs, Pt/RGO etc have also been attempted,8-10 in view of the increased catalytic activity of the composites compared to the pristine material. Some interesting examples are summarized here. The superior electrocatalytic activity of Cu2ZnSn(S0.5Se0.5)4 nanocrystals compared to Pt, resulted in a higher PCE of 3.01% for a QDSC prepared with the multi-element chalcogenide as a CE and TiO2/CdS/CdSe/ZnS as the photoanode.11 A cell with a RGO-Cu2S composite film as a CE and a TiO2/CdS/CdSe/ZnS film as the photoanode significantly outperformed an equivalent cell with Pt as a CE.5 The RGO-Cu2S composite showed a lower charge transfer resistance (Rct), lower overpotential for polysulfide reduction, and a higher exchange current density and thus the corresponding cell exhibited an increased PCE of 4.4% compared to the Pt based cell, which delivered a PCE of 1.6%.5 Isabella et al., have used the Pt, Au, Cu2S obtained by brass treatment and Cu2S deposited on conducting glass via spray as CEs. With TiO2/CdS as the photoanode, they achieved an efficiency of 1.75% for the Cu2S/glass as CE.12 With PbS as a CE and a TiO2/CdSe film as the photoanode, the ensuing cell yielded a PCE of 3.01% compared to a PCE of 0.75% with Pt as the CE.13 The use of PbS as a CE resulted in reduced internal resistances, recombination rates, and concentration gradients in the electrolyte, which enhanced the short circuit current density (Jsc) and the fill factor (FF) in comparison to the cell with Pt as a CE.13 For cells with ZnO/ZnSe/CdSe nanocables as the photoanode, and hierarchical Cu2SnS3- and Cu1.8S- microspheres as CEs, PCEs of 3.65 and 4.06% were attained.14 In another work, metal sulfides (CuS, CoS, NiS, and PbS) were supported on ITO porous films and employed in cells with TiO2/CdS/CdSe/ZnS films as the photoanodes and the highest performance was obtained for a cell with CuS as a CE, with a PCE of 1.47%. 7 Jun et al., studied five CEs: Pt, graphite, carbon soot, Cu2S and rGO, and compared cell performances, but their study was confined to very few CEs and limited characterization.15 Considering the spectacular trajectory of PCE improvements achieved in QDSCs by innovative photoanode design,1,16,17 stems the need to screen a plethora of different types of CEs to identify CEs that are most efficient at playing a complementary role in augmenting the performance of the photoanode. Besides being cheap and easily processable at low temperature, an ideal CE should permit fast electron transfer to the electrolyte, to minimize the reduction overpotential, which in turn reduces back electron transfer from the photoanode to the elec-

Page 2 of 22

trolyte, thus maximizing PCE. While there are reviews on counter electrodes for QDSCs,18-20 but to the best of our knowledge, there are no reports wherein a broad spectrum of different types of CEs have been employed and evaluated (in terms of cell performances) in equivalent QDSCs with same photoanode and the same hole transport layer. Novel CEs are typically studied in isolation.4,21,22 A study of a range of CEs can provide a deeper understanding of how counter electrodes control PCEs of QDSCs. To bridge this gap, here we present an elaborate study on the use of (i) high surface area carbon: in the form of MWCNTs, and SWCNTs supported on FTO, candle soot and Carbon fabric composed of mesh like network of carbon fibers, (ii) a metal sulfide: Cu2S, (iii) a transition metal oxide: WO3 (and variants), (iv) a conducting polymer: PEDOT, as CEs in QDSCs. All the said CEs were used in cells with a TiO2/CdS film as the photoanode and a sodium sulfide solution as an electrolyte. The materials used for CEs were either cheaply procured or prepared economically at lab-scale. Such a study furnished interesting insights into how a CE controls the QDSC performance, which we believe will be useful for researchers working on QDSCs, as it will help in developing efficient CEs suitable for a given QDs sensitized electrode. Future directions for developing high performance CEs are also discussed. Experimental Preparation of TiO2/CdS photoanode and storage Prior to applying TiO2 paste, F:SnO2 coated optically transparent conductive glass substrates (FTO) were cleaned in a soap solution followed by sonication in ethanol and acetone 1:1 v/v mixture by using an ultrasonic bath. Finally, isopropyl alcohol was used for cleaning the substrates. 5 mg of TiO2-P25, (free gift from Evonik) was transferred into a clear solution of acetyl acetone (7.5 mL) + water (2.5 mL), and sonicated for 10 minutes. To this, 1 drop of Triton X-100 surfactant (Sigma Aldrich) was added to prevent TiO2 particle aggregation. Further, this paste was sonicated in a water bath for 10 more minutes, and then the paste was applied onto the cleaned FTO substrates by doctor blading technique. These films were heated to 60 ˚C followed by sintering at 500 ˚C. This first layer of TiO2 is mesoporous, and it is conducive to deposit the CdS QDs. A second layer of TiO2 was applied onto the first layer, to increase the QD loading, and the same annealing procedure was followed. Finally these films were placed in a 40 mM aqueous TiCl4 solution at 70 ˚C for 25 minutes, rinsed in deionized water and sintered at 500 ˚C for another 30 minutes (i) to reduce the surface defects and (ii) for better inter-particle connectivity, so as to minimize the shunt paths. CdS QDs were grown onto the TiO2 surface by the successive ionic layer absorption and reaction (SILAR) technique. To do this, 0.1 M sodium sulfide and 0.1 M cadmium acetate solutions in methanol were taken separately. The TiO2 film was first immersed in the cadmium precursor solution then immersed in the sulfide solution. The QDs growth time was fixed to 2

2

ACS Paragon Plus Environment

Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

minutes. The film was kept for 2 minutes exactly in each precursor solution and followed by rinsing in methanol. After each rinse, the films were dried in a hot air oven for 2 minutes. This was considered as 1 SILAR cycle for CdS QD growth. A 5% aqueous HCl solution was used to remove the CdS QDs which were inadvertently deposited over FTO instead of TiO2. The electrode was subjected to 6 more cycles of SILAR and the TiO2/CdS photoanode was obtained. The TiO2/CdS films were used after storage in dark for 10 days at 45 ˚C. Preparation of CEs and cell fabrication WO3/FTO 1 g of Tungsten metal powder (Sigma Aldrich) was added slowly to a hydrogen peroxide (30%, Merck) (10 mL) solution. The solution was aged for 7 h at room temperature, to allow the exothermic dissolution of the metal power in H2O2. Once the effervescence subsided, the solution was centrifuged and the decant solution was collected. This pale yellow colored solution of polyperoxotungstic acid was drop-cast on FTO substrates and the films were heated at for 100 ˚C for 2 h. The peroxo-groups decompose upon heating to ensue in WO3 films deposited over FTO substrates, which were stored in a desiccator. SWCNTs/FTO SWCNTs procured from Cheap Tubes Inc, were functionalized by using a procedure described in detail, in a previous report.23 1 mg of the functionalized SWCNTs were dispersed in ultrapure water (5 mL) by ultrasonication for 10 minutes. The resulting dispersion of SWCNTs, was used as the medium for electrophoresis. Two pre-cleaned FTO substrates were used as cathode and anode, and immersed in the dispersion of SWCNTs. A constant dc potential of 30 V was applied across the electrodes for three minutes. A black colored deposit was obtained on one of the two FTO substrates. The films were dried for 12 h, in a hot-air oven at 60 ˚C and stored in a vacuum filled desiccator. PEDOT/FTO A clear solution of 0.1 M Lithium trifluoromethanesulfonate or triflate (LiCF3SO3), 1 M camphor 10-sulfonic acid, and 0.1 M ethylenedioxythiophene (EDOT) in isopropyl alcohol (10 mL), all sourced from Sigma-Aldrich, was used as the precursor bath. PEDOT was electropolymerized from the monomer onto a pre-cleaned FTO conducting substrate in a three electrode configuration, by using a Pt rod as the auxiliary electrode, a FTO substrate as the working electrode (WE) and an Ag/AgCl/KCl as the reference electrode under potentiostatic conditions, by applying a constant dc bias of +0.8 V to the working electrode for 3 min. A thick blue colored deposit of PEDOT was obtained over the FTO substrate, and the films were rinsed thoroughly in isopropanol and dried at room temperature in air, and then stored in a desiccator. Cu2S/Brass foil

Cu2S was grown over a brass foil, by using a wellestablished protocol involving the exposure of a brass foil to a polysulfide electrolyte solution, which yields a Cu2S layer grown from the brass substrate.4 A brass foil with 99% purity, was kept immersed in an aqueous concentrated HCl solution at 70 oC for an hour. The foil was then rinsed in ultrapure water, and dried with a hot air gun. It was then immersed in an aqueous polysulfide solution (1 M sulfur (S) + 1 M Na2S) for a minute. The etched portion of brass foil turned black color indicating the formation of cuprous sulfide. The Cu2S/brass film was washed in ultrapure water, and dried in air at room temperature, and then stored in a vacuum desiccator. Soot/FTO A pre-cleaned FTO glass substrate was exposed to the luminous zone of a candle flame for 5 s and a uniform black colored soot deposit was obtained on the conducting surface. The excess deposit outside of the required geometric area was carefully wiped out using a tissue. The soot/FTO film was stored in a desiccator. F-MWCNTs/FTO To a solution of 1:3 v/v H2SO4:HNO3 (each of 6 M strength, 60 mL), MWCNTs (200 mg) were added and the suspension was refluxed at 90 oC for 10 h. After cooling to room temperature, the reaction mixture was diluted with ultrapure water (100 mL) and then washed with ultrapure water until the supernatant liquid showed a neutral pH. Using ashless fast filter paper (Ø: 150 mm), the resulting brown colored solid, referred to as functionalized (F)MWCNTs was collected, dried at room temperature and dispersed in water by sonication for 2 h. F-MWCNTs were electrophoretically deposited onto the FTO substrate in a two electrode cell with two FTO substrates serving as WE and CE. A dc bias of 30 V was applied for 3 min and the FMWCNTs coated substrates were dried at 60 oC in a hotair oven, and then stored in a desiccator. C-Fabric Carbon (C)-Fabric was procured from Alibaba Pvt. Ltd. It was cut into the desired dimensions using a pair of scissors and cleaned with acetone before use. C-Fabric/WO3-x Polyperoxotungstic acid (mentioned in the preparation of WO3 CE) was used as a precursor solution. A three electrode set-up comprising of C-Fabric of desired dimensions as the WE, a Pt rod as the CE, and an Ag/AgCl/KCl as the reference electrode was used. A fixed dc bias of −0.5V was applied for 5 min. to the C-Fabric or FTO (WE) and blue colored films of WO3-x formed over the CFabric or FTO by the cathodic reduction of the polyperoxotungstate species to WO3-x. The photographs of all nine counter electrodes used in this study are shown in Scheme 1, which also gives an illustration of the working of a QDSC. QDSCs were assembled by using TiO2/CdS photoanodes, which were fabricated under exactly similar conditions, to enable a reliable

3

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

comparison of their solar to electrical power conversion efficiencies with the diverse counter electrodes. The area of photoanodes was maintained to lie in the range of 8 to 10 mm2. A cavity was created using a Parafilm on the photoanode, which served as a spacer, and a 0.1 M sodium sulfide in water and methanol (3:7 v/v) solution was dropped in the cavity using a dropper. The TiO2/CdS and the CE films were then sandwiched by placing the CE film over the TiO2/CdS/Parafilm assembly, with their active areas facing each other. Binder clips were used to hold the cell assembly together, and the cells were used for photovoltaic measurements. Characterization techniques Surface morphology of CE materials were examined using a field emission scanning electron microscope (Carl Zeiss Supra 40 FE-SEM). Transmission electron microscopy (TEM) was performed on a Tecnai G2, FEI operating at an accelerating voltage of 300 kV. Samples were dispersed in acetone, and then transferred onto carbon coated copper grids, and the solvent was evaporated at room temperature prior to use. XRD patterns of the CE materials were recorded on a PANalytical, X’PertPRO instrument with Cu-Kα (λ = 1.5406 Å) radiation. A Bruker Senterra dispersive Raman microscope spectrometer, with a 532 nm laser excitation used to measure the vibration modes of the counter electrode materials. Photoluminescence (PL) spectra of the photoactive materials were measured using an assembled system comprising of a M266-I unit (monochromator, with continuum light source with housing, light collection system with toroidal optics and a filter wheel with 5 colored glass filters), coupled with an Avaspec-2048 spectrometer detector to collect the data. A suitable filter was utilized during the measurement. Fluorescence lifetime was deducted by time-correlated single photon counting (TCSPC) method with a Horiba Jobin Yvon data station HUB functioning in the TCSPC mode. Excitation source was a nano LED, Pulsed at 370 nm with a 1MHz repetition rate, the pulse duration was 1.3 ns. Light-scattering Ludox solution (colloidal silica) was used to acquire the instrument response function (prompt). Horiba Jobin Yvon DAS6 fluorescence decay analysis software was used to fit the model function (biexponential decays) to the experimental data, with appropriate correction for the instrument response. Current versus potential (I-V) data of QDSCs were measured using a Newport Oriel 3A solar simulator with a Keithley model 2420 digital source meter. A 450 W Xenon arc lamp was the light source which provided a light intensity of 100 mW cm-2 of Air Mass (AM) 1.5G illumination; the spatial uniformity of irradiance was confirmed by calibrating with a 2 cm × 2 cm Si reference cell and re-affirmed with a Newport power meter. Incident photon to current conversion efficiency (IPCE) versus wavelength data were recorded using a Quantum Efficiency Measurement System, Oriel IQE-200™ capable of measurements compliant to ASTM E1021-06. The light source was a 250 W quartz tungsten halogen lamp, the monochromator path length

Page 4 of 22

was 1/8M and the spot size was 1 mm × 2.5 mm rectangular at focus. Linear sweep voltammograms (LSV) and electrochemical impedance spectra (EIS) were recorded on an Autolab PGSTAT 302N equipped with a frequency response analyser and a NOVA 1.9 software. Results and discussion Some aspects of the photoanode Prior to studying the different CEs, the photoanode was characterized. The structural and optical characteristics of the photoanode are presented in Figure 1. The XRD plot of TiO2 (Figure 1a), shows peaks at 2θ = 25.3o, 37.8o, 48.1o, 53.9o, 55.1 o and 62.7 o which match with the (101), (004), (201), (105), (211) and (204) planes of the tetragonal body centered lattice and anatase structure of TiO2, in accordance with PDF (powder diffraction file) # 894921. The diffractogram of CdS (Figure 1a) shows broad peaks at 2θ = 26.4 o, 43.0 o, and 51.9 o, corresponding to the (111), (220), and (311), reflections of the face centered cubic (fcc) lattice of CdS, as per JCPDF # 890440. The UV-visible absorbance spectrum of TiO2 (Figure 1b) shows that it has a flat and negligible absorption in the visible region and shows an onset of absorbance at 390 nm. The λmax for pristine TiO2 is observed at 320 nm. However, pristine CdS shows a relatively flat absorption profile in the UV-region. Therefore in the TiO2/CdS film, a mixed effect is observed, where the overall absorbance increases in the UV region (due to TiO2), but the distinct peak (as observed in TiO2) is now reduced to a flat but high intensity broad signal (due to CdS), peaking at 304 nm. For the TiO2/CdS film, a broad absorption peak is observed in the visible region, with the onset at 520 nm, indicative of the fact that CdS QDs have diffused into the pores of the TiO2 film and it is this portion of the visible spectrum which (< 520 nm), which is harvested by the QDSC, when this film is employed as a photoanode. From the absorption edges, the optical band gaps (Eg) of TiO2 and CdS were estimated to be 3.2 and 2.38 eV from the relation: Eg = 1240/λ (nm). To study the charge transfer mechanism, the photoluminescence spectra of pristine CdS (deposited on glass) and TiO2/CdS films were recorded at an excitation wavelength of 370 nm (Figure 1c). The PL spectrum of pristine CdS QDs shows a very broad emission, spanning from 450 to 620 nm, with a λmax at 519 nm. The peak was deconvoluted into two components; the first emission peak at 517 nm is primarily attributed to band edge excitation, and the second peak at 574 nm is attributed to the presence of surface defects on QDs. The luminescence of CdS is quenched to ~17% of its’ original magnitude ongoing to the TiO2/CdS film, and this is due to fast transfer of photogenerated electrons from the conduction band (CB) of CdS to the CB of TiO2. This was further confirmed by emission decay analysis. The excited electron life times of CdS alone and along with TiO2 were estimated by fitting the decay plots obtained from the TCSPC method. The decay curves accurately fitted into bi-exponential functions, and the aver-

4

ACS Paragon Plus Environment

Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

age electron lifetimes () were determined and these are listed in Table S1 (supporting information). I = B1 exp(-t/τ1) + B2 exp(-t/τ2)

(1)

= ΣBiτi2/ΣBiτi

(2)

In (1) and (2), B1 and B2 are the amplitude coefficients of the excited electron lifetimes (τ1 and τ2), and t is the time after LED excitation. For CdS QDs, the short lived component of 0.25 ns is assigned to the electron-hole (e--h+) recombination at the band edge and the long lived component of 21.9 ns is attributed to electron trapping in the mid-gap defect states. When CdS QDs are tethered to TiO2, the average electron life time was further reduced, it decreases from 21.6 ns (pristine CdS) to 2.76 ns (TiO2/CdS). The short lived component for TiO2/CdS retains almost the same magnitude, the value is 0.33 ns, and it is the time taken for photoexcited electrons to recombine at the band edge, but the long lived component decreases significantly to 4.48 ns, and this is attributed to the time taken by electrons to cascade into the intra-gap trap states formed at the interface of TiO2 and CdS QDs. Our assignments comply well with those reported by Shalom et al., for TiO2/CdS QDs based films.24 The FE-SEM images of TiO2 and TiO2/CdS films are shown in Figure 1d and e respectively. The TiO2 film is characterized by a granular morphology with globular interlinked particles with voids or pores in-between. The pores show a rather large distribution in size, they vary from 5 to 70 nm in dimensions. When the TiO2 film is exposed to the Cd2+ and S2- solutions, during the growth of CdS by SILAR, the CdS QDs are deposited in the pores of the oxide film, thus ensuring an intimate contact between the oxide and QDs, which is beneficial for fast charge transfer, during solar cell working. The micrograph of the TiO2/CdS film also continues to be granular, for the QDs tend to aggregate, in the absence of a capping agent. The lattice scale TEM image of pristine CdS quantum dots is shown in Figure 2a. The image shows the presence of discrete QDs, approximately 3-6 nm in dimensions. The inter-fringe separation is about 0.29 nm, which matches very well with an inter-planar spacing of 2.91 Å, corresponding to the (200) plane of CdS with a face centered cubic (fcc) lattice (PDF (powder diffraction file): 652887). The TEM image of TiO2/CdS assembly is shown in Figure 2b. The image show mingling nanoparticles of TiO2 with indistinctive shapes, which are decorated with CdS QDs. The grain boundaries are fuzzy for both the QDs and TiO2 at low magnification. The corresponding lattice scale image of the TiO2/CdS assembly (Figure 2c), shows large nanoparticles of TiO2, 20-40 nm in size, represented with the dashed lines. The TiO2 particles are composed of distinct lattice fringes with an inter-fringe distance of 0.36 nm, which is close to a d = 3.51 Å, corresponding to the (101) plane of TiO2 with a body centered tetragonal structure (PDF: 894921). The TiO2 nanoparticles are juxtaposed with the CdS QDs, in some portions,

there is significant overlapping of lattice fringes indicating that the QDs are tethered to the TiO2 scaffold. Discrete CdS QDs are still observed in the film, and these are ensconced in dotted ellipses in the image. The size of the CdS QDs varies from 3 to 7 nm. The selected area electron diffraction (SAED) pattern of the TiO2/CdS assembly (Figure 2d) shows a diffused ring pattern superimposed with bright and dull spots. The spots were indexed to (101), (103), (200) and (116) planes of the tetragonal structure of TiO2, for these spots matched with d = 3.51, 2.42, 1.90, and 1.37 Å. Bright spots from the crystalline CdS QDs were indexed to the (200), (220), (311), (400), (331) and (440) planes of the fcc structure. The d-values for these planes were 2.9, 2.05, 1.74, 1.44, 1.34, and 1.04 Å. The coexistence of spots from TiO2 and CdS in the SAED pattern confirms a good mixing of the two components. Counter electrode characterization Raman spectroscopic analysis In this study, the films which were employed for use as CEs in QDSCs, are: F-MWCNTs, soot, SWCNTs, C-Fabric, WO3, WO3-x supported on C-Fabric, Cu2S, and PEDOT. The Raman spectra of the CEs are presented in Figure 3. The molecular vibrations of the carbonaceous materials are compared in Figure 3a, and this has been accomplished by comparing their ID/IG values. The D-band in carbon based materials arises from the out-of-plane vibrations or ring breathing modes of carbon atoms due to structural defects (atomic vacancies or presence of oxygen containing functional groups along the carbon framework) or is the measure of deviation from the perfect graphitic structure. The G-band originates from the in-plane vibrational modes of sp2 hybridized carbon atoms. The Dand G- band positions, and the corresponding ID/IG ratios are provided in Table 1. In the table, the carbon based materials are arranged in the increasing order of the proportion of defects. Among all, the ID/IG ratio is the highest for the C-Fabric, it is 2.28, and this high concentration of defects in C-Fabric possibly forms during the synthetic process, for large mechanical forces are applied to yield lengthy C-fibers, and this is responsible for the increased defects on its surface. Despite this defect concentration, a high PCE is achieved for the cells containing C-Fabric or C-Fabric/WO3-x as CEs, for this disadvantage is offset by (i) its’ extremely low sheet resistance, 8 to 15 Ω cm-2, lower than that of pristine FTO (30 Ω cm-2), which serves as the current collector for the bulk of the CE materials used in this study and (ii) the high effective surface (catalytic) area offered by the mesh like structure of the intertwined C-fibers in the C-Fabric. The ID/IG ratio is surprisingly the least for soot, and comparable for SWCNTs, which have slightly higher values, in that order. The low ID/IG ratio of 1.12, obtained for soot is ascribed to the high preparation temperature of 1000 oC, which perhaps results in the formation of interlinked carbon particles, with minimum defects and without the formation of any oxygenated functionalities.

5

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Raman spectrum of Cu2S (Figure 3b) shows a narrow intense peak at 470 cm-1, which is assigned to the vibrational (stretching) mode of the covalent bond existing between two sulfur atoms, and a broad peak at about 290 cm-1 is attributed to the Cu–S bond vibration.25 For WO3, the bands at 267 and 322 cm-1 correspond to the O– W–O bending modes of the bridging oxygen and the peaks at 709 and 809 cm-1 represent symmetric and asymmetric O–W–O stretching vibrations. These peaks are very close to the wavenumbers of the four strongest modes of crystalline WO3.26,27 The spectrum of WO3-x, did not show any peaks, probably because of high amorphicity, and therefore it is not shown here. In the case of PEDOT, Raman peaks at 1488, 1426 (strong), 1360, and 1256 cm-1 were observed and these are assigned to the asymmetric (C=C) stretching, symmetric Cα=Cβ(-O) stretching, Cβ-Cβ stretching and Cα-Cα′ (inter-ring) stretching vibrational modes.28,29 Additional medium intensity peaks observed at 988, 576, and 436 cm-1 are attributed to the oxyethylene ring deformation modes.28,29 No peak was observed around 1700 cm-1, which indicated that monomer is not over-oxidized, during PEDOT formation. XRD studies XRD patterns of CE materials are displayed in Figure 4. C-Fabric, C-Fabric/WO3-x, F-MWCNTs, SWCNTs, and soot, show characteristic broad humps at 2θ = 24.7o corresponding to the interplanar d-spacing of 3.60 Å (Figure 4a). The latter peak is attributed to the (002) plane of graphite, in concurrence with PDF # 75-1621. The interlayer spacing between the graphitic sheets is enlarged in the other carbon nanostructures, possibly due to the presence of functional groups. SWCNTs are characterized by an additional peak at 2θ = 12.0o, corresponding to d = 7.36 Å, which agrees with d = 8.8 Å, obtained for GO by Gao et al., in a previous report.30 Authors attributed this peak to the enhanced separation between graphitic layers, induced by the presence of bulky function groups in between the layers. In another report, carbon nanostructures with 2θ peaks < 15o,31 were associated with interlayer separation corresponding to the (001) plane and therefore this attribute is observed in SWCNTs here. CFabric also shows a very broad peak at 2θ = 44.02o, corresponding to d = 2.05 Å, and it represents the C-C atomic separation in the carbon nanostructure. For the CFabric/WO3-x composite, the contributions from WO3-x superimposed on the contributions from the underlying C-Fabric, so reliable assignments (from carbon) could not be made. Cu2S produced a pattern with peaks at 2θ = 49.5o, 72.5o and 87.4o, and these correspond to d = 1.85, 1.31 and 1.12 Å, which align with the (110), (114) and (220) planes of the hexagonal (primitive) structure, based on the PDF # 892670 (Figure 4b).32 Reduced WO3 or WO3-x is almost amorphous, for it shows three extremely broad

Page 6 of 22

peaks at 2θ = 24.7o, 34.9o and 56.8o, which correspond to d = 3.56, 2.57 and 1.62 Å, and these match with the (012), (122) and (412) planes of orthorhombic WO3 (PDF # 894479). WO3 is polycrystalline, for it shows multiple peaks in its’ XRD pattern, and prominent peaks are observed at 2θ = 23.64o, 28.53o, 33.84 o, 41.76 o, 50.06 o, 55.34 o, 61.7 o and 76.5 o, and these correspond to d = 3.75, 3.13, 2.64, 2.16, 1.82, 1.66, 1.50 and 1.24 Å, which match with the (020), (112) (220) (222) (322) (204) (134) and (235) planes of the orthorhombic crystal structure of WO3, as per PDF # 894479. Electron microscopy studies FE-SEM images of the CE materials are shown in Figure 5 and 6. The images of F-MWCNTs (Figure 5a and b) show a mesh like structure composed of the functionalized CNTs; while the fibrillar structures are retained but the discrete character of the CNTs is not visible. Further, a few particulate structures appear to be superimposed over the tubes, in some regions, indicating that the tubes get ruptured to accommodate the oxygen containing functional groups (i.e., C=O, C-O-C, C-OH etc), during acid treatment. The micrograph of candle soot (Figure 5c) reveals a dense particulate morphology, with a few pores interspersed in between. The particles have no unique shape, and they are completely aggregated along multiple length scales. Since the soot was directly adsorbed from the candle flame to the FTO substrate, the soot deposit has a wavy texture, it appears to be thicker along the crests and thinner along the troughs. Pristine C-Fabric has a woven mat like structure (Figure 5d and e); it is made up of carbon fibers which are neatly arranged into a webbing that imparts the netting like appearance to the C-Fabric, as a whole. The image of SWCNTs (Figure 5f) reveals the nanotubes to be well-dispersed, despite the chemical modification, and there are no carbonaceous particles seen to be flanking the tubes, indicating that the single walled structures are not severely impacted by the acid treatment and sonication. The morphology of Cu2S (Figure 6a and b) shows flaky petal like shapes which are uniformly distributed across the surface of the brass foil, and they appear to be perpendicularly oriented with respect to the brass foil. Although there is some degree of agglomeration of the flaky shapes, still, they seem to be discrete enough to offer a high catalytic area for the reduction of the oxidized polysulfide ions (from the electrolyte), during solar cell operation, when the Cu2S/brass film is employed as a CE therein. The morphology of WO3 (Figure 6c), is almost featureless, for the image shows the oxide film to be made up of a compactly packed network of particles of indistinctive shapes, with hardly any pores separating them. The micrograph of WO3-x (Figure 6d) comprises of a connected micron-sized particles, separated by pores. The micrograph of C-Fabric/WO3-x (Figure 6e) shows the fibers of C-fabric to be coated with particles of WO3-x; the thick

6

ACS Paragon Plus Environment

Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

deposits of WO3-x tethered to the C-Fabric has two distinctive attributes. (i) Due to the reduced nature of the oxide (in WO3-x), the electronic conductivity is expected to higher than that of WO3/FTO, and (ii) unlike the planar FTO, the interlaced structure of C-Fabric, on which the WO3-x particles attach to, have a higher surface area. As a consequence, the solar cell performance of the cell with C-Fabric/WO3-x as the CE is expected to be better than that of cells with FTO/WO3 or FTO or C-Fabric as the CEs. The image of PEDOT (Figure 6f) reveals interconnected particles of cauliflower like shapes, with large pores in-between. This morphology is typical of electropolymerized PEDOT films.33 J-V characteristics and spectral response QDSCs were assembled with TiO2/CdS films as the photoanodes, and nine different CEs. A solution of 0.1 M Na2S was used as the electrolyte in all the cells. The current density (J) versus voltage (V) plots for the QDSCs are presented in Figure 7a and b. The solar cell parameters: short circuit current density (JSC), open circuit voltage (VOC), fill factor (FF) and PCE (η) for all the cells are summarized in Table 2. The average values of PCEs are also included in the Table 2; for each cell configuration, five cells were considered. The average parameters are furnished in Table S2 (supporting information). The magnitudes of all the solar cell parameters are not only governed by the photoanode and the electrolyte, but are also strongly dependent of the “counter electrode”. Across the entire spectrum of cells studied herein, the Voc is found to vary between 0.72 to 1.07 V. The lowest Voc is obtained for a cell with Cu2S as the CE. The highest VOC is achieved for the cell with WO3 as the CE. The VOC of the cell with CFabric/WO3-x as the CE, comes a close second (1.04 V). All the cells with carbonaceous CEs exhibit reasonably high values of VOC, in the range of ~0.80 to 0.96 V. On comparing the JSC values achieved for the cells, the highest JSC was observed for the cell with Cu2S/brass as the CE (12.56 mA cm-2). This performance is due to the high catalytic area offered by the Cu2S petal (flake) like shapes, which enables efficient reduction of polysulfide species, and thus leads to more charge generation. The cell containing CFabric shows the second highest JSC (10.04 mA cm-2), and this again is due to its low electrical resistance. The cells with other electrodes show JSC values that lie in the range of 6.5 to 10 mA cm-2. Besides VOC and JSC, a high FF also contributes to maximizing PCE of a QDSC. A high FF is generally obtained for a cell, if the shunt resistance is high and the series resistance is low. The highest FF is delivered by the cell with C-Fabric/WO3-x as the CE (~ 51.5), indicating that the series resistance is minimum and the shunt resistance is maximum, among all cells here. This high FF translates to the highest PCE of 4.6% for the champion cell, among all cells studied in this report. The cell with C-Fabric delivers a PCE of 3.97%, closely followed by the cell with F-MWCNTs as the CE (3.78%). By examining all the PCEs in Table 2, it is evident that car-

bonaceous structures as CEs are very effective in producing cells with high efficiencies. The C-Fabric/WO3-x electrode outperforms all other CEs, due to the synergy between the characteristics of C-Fabric and WO3-x: (i) low sheet resistance of C-Fabric, and its’ high surface area due to the netting like morphology, which allows high WO3-x loading during electrodeposition, and (ii) the good catalytic activity of WO3-x, and its high electrical conductivity that facilitate electron transfer to the electrolyte, during cell operation. We also ran a comparison with literature values. Radich et al., fabricated a Cu2S/rGO composite and used it as a CE with CdS/CdSe QDs photoanode in a QDSC and achieved a PCE of 4.4%.5 Deng et al., screen-printed Cu2S QDs and conductive carbon on FTO and when it was employed as a CE in a CdS/CdSe based QDSC, the PCE of the device was 3.71%.34 Zhao et al., compared cell performances with Cu2S/FTO and Cu2S/brass and Pt as CEs. They obtained PCEs of 5.21, 5.41 and 1.68% respectively.4 The use of carbon cloth as a CE is reported in dye sensitized solar cells (DSSCs), and not in QDSCs, by other researchers. Tathavadekar et al., fabricated a conducting carbon cloth from a cellulose fabric, from a complex pyrolysis route, and then obtained PCEs of 5.8 and 7% for DSSCs with C-cloth and Pt/FTO as CEs.35 The perfromance is reversed in our study, for we found that in QDSCs, with C-Fabric, the PCE was higher than that achieved with Pt. Similarly WO3 has also not been used in QDSCs, but has beeen used in DSSCs. Li et al., used NH3-treated WO3 as a CE in a DSSC, and it delivered a PCE of 5.9%, which was comparable to the PCE of the cell with Pt as CE (6.0%).36 Here, we observed that with WO3/FTO and WO3-x/FTO as CEs in QDSCs, the PCEs are higher, 2.1 and 3.66%, compared to that obtained with Pt as the CE (1.69%), reiterating that other catalytic materials are better performers compared to Pt in QDSCs, thus contrasting the trends observed in DSSCs. With carbon-soot as CE and TiO2/CdS as the photoanode, authors in,15 reported a PCE of 1.16%, which is lower than the value obtained here with soot (2.95%). For a cell with MWCNTs as the CE, and TiO2/CdSe/ZnS as the photoanode, a PCE of 2.39% was obtained.37 Yeh et al., compared the performances of QDSCs with three conducting polymers as CEs, namely, poly(thiophene) (PT), poly(pyrrole) (PPy), and PEDOT, and achieved PCEs of 0.09, 0.41 and 1.16% respectively.38 In the present work, the PCE of the champion cell with PEDOT as the CE is 2.79%. A more detailed literature survey on solar cell parameters is provided in Table S3 (supporting information). The spectral response or IPCE versus wavelength curves of QDSCs with different CEs are shown in Figure 6c and d. The IPCE values are the highest in the 320-520 nm wavelength range for the cell with C-Fabric/WO3-x as CE, which is in line with the highest PE achieved for this cell.

7

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The maximum IPCE exhibited by the cell with the CE: CFabric/WO3-x, is 82% at 430 nm. The response is comparable for the cells with F-MWCNTs and C-Fabric as the CEs, with slight variations, which also matches with the minor differences in their PCEs. The cells with other electrodes show lower values of IPCE in comparison with the cells with the aforesaid CEs. As expected, the cell with SWCNTs as the CE shows the lowest values of IPCEs in the visible region. Redox potential and EIS studies For fast electron transfer to the polysulfide radical in the electrolyte at the counter electrode, the CE should ideally offer a low overpotential. The overpotential (η) required to drive the reaction at a certain current density J, gives rise to a charge transfer resistance (RCT = η/J), which acts as a series resistance in the solar cell.39 Inorder to evaluate the electrocatalytic activities of the CE films for the reduction of polysulfide, LSV plots were recorded for the CE films (serving as working electrodes) in a 0.02 M Na2S in methanol/water solution, containing 0.2 M KCl as the supporting electrolyte, with Pt as the auxiliary electrode. The potential is volts was also converted to energy in eV. The standard redox potential of Sn2-/S2- is −0.6 V (versus normal hydrogen electrode or NHE), which is −3.9 eV versus vacuum. The LSV curves of the various CEs are presented in Figure 7a and b, and all the curves show a broad reduction peaks in the −3.7 to −4.3 eV. CEs such as C-Fabric/WO3-x, C-Fabric and WO3-x show the least overpotentials, as their redox potentials are very close to the standard reduction potential of Sn2-/S2- redox couple. A high magnitude of reductive peak current density is also indicative of a high electrocatalytic activity. Table 3 furnishes the peak current densities and the redox potentials of the CEs. The C-Fabric/WO3-x electrode, shows the least overpotential and the highest peak current density (during reduction), which is responsible for the highest PCE. Further, tungsten is earth-abundant, tungsten oxide films are easily processable at low temperatures, can be fabricated with high uniformity over large areas using low cost procedures and are chemically stable as well- aspects which are advantageous in terms of long term operation, and scale-up for developing large area QDSCs. Since oxygen vacancy is a common feature of many metal oxides, the findings in this work can be extrapolated to other metal oxides (such as MoO3-x) and thus offering the possibility for developing composites with them. To quantify the charge transfer resistance at the CE/electrolyte interface, a key factor that controls QDSC performance, EIS plots were recorded for QDSCs with the different CEs, under a white light irradiance of 20 mW cm-2, over a frequency range of 1 MHz to 0.1 Hz, and under an ac amplitude of 20 mV. All the plots, shown in Figure 7(c-f) contain a skewed arc, extending into a broad semicircle, ongoing from high to low frequency. The first

Page 8 of 22

arc corresponds to the electron transfer between the polysulfide electrolyte and the CE interface, and the second semi-circle represents the electron transport in the QD sensitized TiO2 layer and the recombination process at the QD-sensitized TiO2 and electrolyte interface. For the cells with the following configuration: TiO2/CdS-Sn2-/S2-CE, the first semicircle in the corresponding Nyquist plot gave the charge transfer resistance (RCT) at the CE/S2- interface. The RCT magnitudes were deduced to be: 247, 202, 160, 152, 125, 110, 45, 38, and 7 Ω cm2 corresponding to the SWCNTs, WO3, PEDOT, Cu2S, soot, WO3-x, F-MWCNTs, C-Fabric and C-Fabric/WO3-x electrodes. The magnitude of RCT was observed to be the least for the C-Fabric/WO3-x electrode, suggesting that electron transfer to the electrolyte from this CE is facile. C-Fabric and F-MWCNTs also moderately low RCT values. Superior solar cell performances of the cells based on these electrodes, is also credited to their low RCT values. SWCNTs shows the highest RCT, which again, is another reason for the least PCE. Future directions Considering the following pre-requisites: high electrocatalytic activity, good chemical- and photo- stability, high electronic conductivity, cost-effectiveness, and nontoxicity, there are many exciting possibilities the development of yet unexplored CEs for use in QDSCs. Electropolymerized films of conducting polymers doped with carbon nanostructures (as dopants) can be attempted. These include yet unexplored composite films of poly(methylpyrrole) or PMP doped with SWCNTs prefunctionalized with carboxylic groups (COO-), where PMP is a robust, cheap, chemically stable, room temperature processable (in film form), and a highly scalable conducting polymer, and functionalized SWCNTs are good electronic conductors, and therefore the synergy between the properties of these two materials will come to the fore in the PMP/SWCNTs film, thus rendering it suitable for use as a CE. The proportion of SWCNTs required for making such composite films is extremely low, and the monomer baths are highly reusable (when stored at ~3-5 oC) thus making this process economically viable as well. Other examples include composites of conducting polymers with metal sulfides, such as poly(indole-5-carboxylic acid) (PICA) with Cu2S, wherein the metal sulfide will exhibit high electrocatalytic activity, and the conducting polymer will show good electrical conductivity. This composite can be prepared by simply electropolymerizing the monomer (indole-5-carboxylic acid), in the presence of pre-synthesized Cu2S nanoparticles. The ensuing PICA/Cu2S composite films are again easy to fabricate, up-scalable to large areas if required, and their compositions can be optimized for CE application. The advantage of using composites with conducting polymers is the following: conjugated polymers can be easily fabricated in the form of uniform thin films over FTO substrates, and they are generally compatible with most of the carbon nanostructures such as reduced graphene oxide and CNTs, metal sulfides (NiS, Cu2S, CuS, Cu-Zn-Se-S

8

ACS Paragon Plus Environment

Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

etc), metal oxides (such as MoO3-x) and even metal nanoparticles (like Au, Pt etc) and they are highly stable in sulfide electrolyte. Therefore composite formation to improve the catalytic activity of the conducting polymer with any of the above is facile, and is expected to lead to cells with high performance. Some more possibilities can be: poly(3,4-ethylenedioxypyrrole) (PEDOP) with MWCNTs, PICA with RGO, PMP with NiS or CuS or MoO3-x and so forth. These directions can be adopted to develop many such novel composite films, where the synergy between the two components can ameliorate the QDSC performance, and which have not been applied to QDSCs till date.

Email: [email protected], Tel: +91-40-23016024, Fax: +9140-23016003. Acknowledgments Financial support from the Solar Energy Research Initiative-Department of Science & Technology (DST/TM/SERI/2K12-11(G)) is gratefully acknowledged. PNK is thankful to CSIR for senior research fellowship. REFERENCES (1)

(2)

Conclusions The role of CE in controlling the performance of a QDSC has been demonstrated in this report, by using nine different CEs: SWCNTs, WO3, PEDOT, Cu2S, candle soot, F- MWCNTs, WO3-x, C-Fabric, and C-Fabric/WO3-x. QDSCs were constructed with TiO2/CdS photoanodes and a polysulfide electrolyte. In a QDSC, the efficacy of a CE is a direct measure of the effectiveness of the reduction process of the polysulfide species at the CE/electrolyte interface, which then regenerates the QDs. How effective the CEs were in catalyzing this reduction was reflected in the PCEs of the cells: 2.02, 2.1, 2.79, 2.88, 2.95, 3.78, 3.66, 3.96, and 4.6 % respectively, based on the aforesaid CEs. The variation of IPCEs with wavelength for the cells was observed to concur with the PCE trend. Among all the CEs, the C-Fabric/WO3-x based cell exhibited the highest PCE, and this was attributed to the least polysulfide reduction potential, a reasonably good electrocatalytic activity (evidenced from the highest reduction peak current density), and a low sheet resistance of the C-Fabric and its’ mesh like morphology that maximized WO3-x loading. Sole WO3-x based cell showed a high FF and a VOC > 1 V, and this aided in improving the FF of the cell with the CFabric/WO3-x CE. Most of the carbon nanostructures based and Cu2S based CEs yielded moderately high PCEs, when used in QDSCs. This study showed that WO3 based electrodes, particularly oxygen deficient WO3-x, can serve as efficient CEs in QDSCs. Concepts to design novel conducting polymer composites based CEs are also discussed, and we expect this report to be of great significance for developing QDSCs with high efficiencies.

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

Supporting Information Tables of emission decay parameters and solar cell parameters (from this work and from literature). This material is available free of charge via the Internet at http://pubs.acs.org.

(12)

Corresponding Author Melepurath Deepa

(13)

Department of Chemistry, Indian Institute of Technology Hyderabad, Kandi-502285, Sangareddy, Telangana (India).

(14)

Du, J.; Du, Z.; Hu, J. S.; Pan, Z.; Shen, Q.; Sun, J.; Long, D.; Dong, H.; Sun, L.; Zhong, X.; Wan, L.-J. Zn-Cu-In-Se Quantum Dot Solar Cells with a Certified Power Conversion Efficiency of 11.6%. J. Am. Chem. Soc. 2016, 138 (12), 4201–4209. Ren, Z.; Wang, J.; Pan, Z.; Zhao, K.; Zhang, H.; Li, Y.; Zhao, Y.; Mora-Sero, I.; Bisquert, J.; Zhong, X. Amorphous TiO2 Buffer Layer Boosts Efficiency of Quantum Dot Sensitized Solar Cells to over 9%. Chem. Mater. 2015, 27 (24), 8398– 8405. Chen, H.; Zhu, L.; Liu, H.; Li, W. Efficient Iron Sulfide Counter Electrode for Quantum Dots-Sensitized Solar Cells. J. Power Sources 2014, 245, 406–410. Zhao, K.; Yu, H.; Zhang, H.; Zhong, X. Electroplating Cuprous Sulfide Counter Electrode for High-Efficiency Long-Term Stability Quantum Dot Sensitized Solar Cells. J. Phys. Chem. C 2014, 118, 5683–5690. Radich, J. G.; Dwyer, R.; Kamat, P. V. Cu2S Reduced Graphene Oxide Composite for High-Efficiency Quantum Dot Solar Cells. Overcoming the Redox Limitations of S2 /Sn2- at the Counter Electrode. J. Phys. Chem. Lett. 2011, 2, 2453–2460. Shen, C.; Sun, L.; Koh, Z. Y.; Wang, Q. Cuprous Sulfide Counter Electrodes Prepared by Ion Exchange for HighEfficiency Quantum Dot- Sensitized Solar Cells. J. Mater. Chem. A 2014, 2, 2807–2813. Chen, H.; Zhu, L.; Liu, H.; Li, W. ITO Porous FilmSupported Metal Sulfides Counter Electrodes for High Performance Quantum Dots-Sensitized Solar Cells. J. Phys. Chem. C 2013, 117, 3739–3746. Santra, P. K.; Kamat, P. V. Tandem Layered Quantum Dot Solar Cells . Tuning the Photovoltaic Response with Luminescent Ternary Cadmium Chalcogenides. J. Am. Chem. Soc. 2012, 135, 877–885. Kaniyoor, A.; Ramaprabhu, S. Gold Nanoparticle Decorated Multi-Walled Carbon Nanotubes as Counter Electrode for Dye Sensitized Solar Cells. J. Nanosci. Nanotechnol. 2012, pp 8323–8329. Dao, V. D.; Hoa, N. T. Q.; Larina, L. L.; Lee, J. K.; Choi, H. S. Graphene-Platinum Nanohybrid as a Robust and Low-Cost Counter Electrode for Dye-Sensitized Solar Cells. Nanoscale 2013, 5 (24), 12237–12244. Cao, Y.; Xiao, Y.; Jung, J. Y.; Um, H. D.; Jee, S. W.; Choi, H. M.; Bang, J. H.; Lee, J. H. Highly Electrocatalytic CuZnSn(S1xSex)₄ Counter Electrodes for Quantum-Dot-Sensitized Solar Cells. ACS Appl. Mater. Interfaces 2013, 5 (3), 479–484. Milan, R.; Hassan, M.; Selopal, G. S.; Borgese, L.; Natile, M. M.; Depero, L. E.; Sberveglieri, G.; Concina, I. A Player Often Neglected: Electrochemical Comprehensive Analysis of Counter Electrodes for Quantum Dot Solar Cells. ACS Appl. Mater. Interfaces 2016, 8 (12), 7766–7776. Tachan, Z.; Shalom, M.; Hod, I.; Rühle, S.; Tirosh, S.; Zaban, A. PbS as a Highly Catalytic Counter Electrode for Polysulfide Based Quantum Dot Solar Cells. J. Phys. Chem. C 2011, 115, 6162–6166. Xu, J.; Yang, X.; Wong, T. L.; Lee, C. S. Large-Scale Synthesis of Cu2SnS3 and Cu1.8S Hierarchical Microspheres

9

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

as Efficient Counter Electrode Materials for Quantum Dot Sensitized Solar Cells. Nanoscale 2012, 4 (20), 6537–6542. Jun, H. K.; Careem, M. A.; Arof, A. K. Performances of Some Low-Cost Counter Electrode Materials in CdS and CdSe Quantum Dot-Sensitized Solar Cells. Nanoscale Res. Lett. 2014, 9 (1), 69–76. Ren, Z.; Wang, J.; Pan, Z.; Zhao, K.; Zhang, H.; Li, Y.; Zhao, Y. Amorphous TiO2 Buffer Layer Boosts Efficiency of Quantum Dot Sensitized Solar Cells to over 9%. Chem. Mater. 2015, 27, 8398–8405. Gong, X.; Wang, J.; Bonn, M.; Bisquert, J.; Zhong, X. Boosting Power Conversion Efficiencies of Quantum-DotSensitized Solar Cells Beyond 8% by Recombination Control. J. Am. Chem. Soc. 2015, 137, 5602–5609. Meng, K.; Chen, G.; Thampi, K. R. Metal Chalcogenides as Counter Electrode Materials in Quantum Dot Sensitized Solar Cells: A Perspective. J. Mater. Chem. A 2015, 3, 23074– 23089. Hwang, I.; Yong, K. Counter Electrodes for Quantum-DotSensitized Solar Cells. ChemElectroChem 2015, 2 (5), 634– 653. Hod, I.; Zaban, A. Materials and Interfaces in Quantum Dot Sensitized Solar Cells: Challenges, Advances and Prospects. Langmuir 2014, 30 (25), 7264–7273. Yuan, H. L.; Lu, J. F.; Xu, X. B.; Huang, D. K.; Chen, W.; Shen, Y.; Wang, M. K. Electrochemically Deposited CoS Films as Counter Electrodes for Efficient Quantum DotSensitized Solar Cells. J. Electrochem. Soc. 2013, 160 (9), H624–H629. Kumar, P. N.; Deepa, M.; Srivastava, A. K. Ag Plasmonic Nanostructures and a Novel Gel Electrolyte in a High Efficiency TiO2/CdS Solar Cell. Phys. Chem. Chem. Phys. 2015, 17 (15), 10040–10052. Worsley, K. A.; Kalinina, I.; Bekyarova, E.; Haddon, R. C. Functionalization and Dissolution of Nitric Acid Treated Single-Walled Carbon Nanotubes. J. Am. Chem. Soc. 2009, 131 (50), 18153–18158. Shalom, M.; Dor, S.; Ru, S.; Grinis, L.; Zaban, A. Core / CdS Quantum Dot / Shell Mesoporous Solar Cells with Improved Stability and Efficiency Using an Amorphous TiO2 Coating. J. Phys. Chem. C 2009, 113, 3895–3898. Milekhin, A. G.; Yeryukov, N. A.; Sveshnikova, L. L.; Duda, T. A.; Rodyakina, E. E.; Gridchin, V. A.; Sheremet, E. S.; Zahn, D. R. T. Combination of Surface- and InterferenceEnhanced Raman Scattering by CuS Nanocrystals on Nanopatterned Au Structures. Beilstein J. Nanotechnol. 2015, 6, 749–754. Santato, C.; Odziemkowski, M.; Ulmann, M.; Augustynski, J. Crystallographically Oriented Mesoporous WO3 Films : Synthesis , Characterization , and Applications. J. Am. Chem. Soc. 2001, 123, 10639–10649. White, C. M.; Jang, S.; Lee, H.; Pankow, J.; Dillon, A. C. Photocatalytic Activity and Photoelectrochemical Property of Nano-WO3 Powders Made by Hot-Wire Chemical Vapor

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

(37)

(38)

(39)

Page 10 of 22

Deposition. Electrochem. Solid-State Lett. 2010, 120–122. Garreau, S.; Louarn, G.; Buisson, J. P.; Froyer, G.; Lefrant, S. In Situ Spectroelectrochemical Raman Studies of Poly ( 3 , 4-Ethylenedioxythiophene ) ( PEDT ). Macromolecules 1999, 32, 6807–6812. Xia, Y.; Zhang, H.; Ouyang, J. Highly Conductive PEDOT : PSS Films Prepared through a Treatment with Zwitterions and Their Application in Polymer Photovoltaic Cells. J. Mater. Chem. 2010, 20, 9740–9747. Gao, W.; Alemany, L. B.; Ci, L.; Ajayan, P. M. New Insights into the Structure and Reduction of Graphite Oxide. Nat. Chem. 2009, 1 (5), 403–408. Li, Y. S.; Liao, J. L.; Wang, S. Y.; Chiang, W. H. Intercalation-assisted longitudinal unzipping of carbon nanotubes for green and scalable synthesis of graphene nanoribbons. Scientific Reports 2016, 6, 22755. Wang, S.; Yang, S. Growth of Crystalline Cu2S Nanowire Arrays on Copper Surface : Effect of Copper Surface Structure , Reagent Gas Composition , and Reaction Temperature. chem. mater. 2001, 13, 4794–4799. Bhandari, S.; Deepa, M.; Srivastava, A. K.; Kant, R. PostPolymerization Functionalization of poly(3,4Ethylenedioxythiophene) Films by 1-Fluoro-2-Nitro-4Azidobenzene: Electrochromism and Redox Behavior. J. Mater. Chem. 2009, 19 (16), 2336–2348. Deng, M.; Huang, S.; Zhang, Q.; Li, D.; Luo, Y.; Shen, Q.; Toyoda, T.; Meng, Q. Screen-Printed Cu2S-Based Counter Electrode for Quantum-Dot-Sensitized Solar Cell. Chem. Lett. 2010, 39 (84), 1168–1170. Tathavadekar, M.; Biswal, M.; Agarkar, S.; Giribabu, L.; Ogale, S. Electronically and Catalytically Functional Carbon Cloth as a Permeable and Flexible Counter Electrode for Dye Sensitized Solar Cell. Electrochim. Acta 2014, 123, 248– 253. Song, D.; Chen, Z.; Cui, P.; Li, M.; Zhao, X.; Li, Y.; Chu, L. NH3-Treated WO3 as Low-Cost and Efficient Counter Electrode for Dye-Sensitized Solar Cells. Nanoscale Res. Lett. 2015, 10 (1), 1–6. Zeng, X.; Xiong, D.; Zhang, W.; Ming, L.; Xu, Z.; Huang, Z.; Wang, M.; Chen, W.; Cheng, Y. B. Spray Deposition of Water-Soluble Multiwall Carbon Nanotube and Cu2ZnSnSe4 Nanoparticle Composites as Highly Efficient Counter Electrodes in a Quantum Dot-Sensitized Solar Cell System. Nanoscale 2013, 5 (15), 6992–6998. Yeh, M. H.; Lee, C. P.; Chou, C. Y.; Lin, L. Y.; Wei, H. Y.; Chu, C. W.; Vittal, R.; Ho, K. C. Conducting Polymer-Based Counter Electrode for a Quantum-Dot-Sensitized Solar Cell (QDSSC) with a Polysulfide Electrolyte. Electrochim. Acta 2011, 57 (1), 277–284. Boschloo, G.; Hagfeldt, A. Characteristics of the Iodide / Triiodide Redox Mediator in Dye-Sensitized Solar Cells. Acc. Chem. Res. 2009, 42 (11), 1819–1826.

FIGURES

10

ACS Paragon Plus Environment

Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 1 (a) XRD patterns of TiO2 and CdS and (b) absorption spectra of TiO2, CdS, and TiO2/CdS. (c) Fluorescence spectra of CdS/glass (deconvoluted, represented by solid lines) and TiO2/CdS (λex = 370 nm) and (d) emission decay plots of CdS/glass and TiO2/CdS (λex = 370 nm and λem = 515 nm). FE-SEM images of (e) TiO2 and (f) TiO2/CdS films.

11

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

Figure 2 (a) Lattice scale image of pristine CdS QDs, (b) TEM image-, (c) lattice scale image- and (d) SAED pattern- of a TiO2/CdS assembly.

12

ACS Paragon Plus Environment

Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 3 Raman spectra of (a) soot, C-fabric, SWCNTs, and F-MWCNTs, and (b) Cu2S, WO3 and PEDOT.

13

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

Figure 4 XRD patterns of (a) C-Fabric, C-Fabric/WO3-x, F-MWCNTs, SWCNTs, and soot, and (b) Cu2S, PEDOT, WO3-x and WO3.

14

ACS Paragon Plus Environment

Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 5 FE-SEM images of (a and b) F-MWCNTs, (c) soot, (d and e) C-Fabric, and (f) SWCNTs.

15

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

Figure 6 FE-SEM images of (a and b) Cu2S, (c) WO3, (d) WO3-x, (e) C-Fabric/WO3-x, with inset showing WO3-x coated C-fibers and (f) PEDOT.

16

ACS Paragon Plus Environment

Page 17 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

-2

Figure 7 (a and b) J-V characteristics (recorded under 1 sun irradiance (100 mW cm )) and (c and d) IPCE versus wavelength plots of QDSCs with TiO2/CdS films as the photoanodes, 0.1 M Na2S solutions as electrolytes, and with different counter electrode films.

17

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 22

Figure 8 (a and b) Linear sweep voltammograms of different CEs (as working electrodes) in a 0.02 M Na2S solution in an ultrapure water:methanol solution (3:7 v/v) containing 0.2 M KCl as supporting electrolyte. A Pt rod was used as the counter electrode and an Ag/AgCl/KCl was used as a reference electrode. The potential in volts on the abscissa was converted to eV, by using + the relation (E (eV) = −(x V (vs. Ag/Ag ) + 0.197 V + 4.5)). (c-f) Nyquist plots of QDSCs with TiO2/CdS films as the photoanodes, 0.1 M Na2S solutions as electrolytes, and with different counter electrode films.

18

ACS Paragon Plus Environment

Page 19 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

SCHEME

Scheme 1 (a) Cartoon illustrating the working of a QDSC, and (b) a photograph showing all counter electrodes used in QDSCs. In (b), the C-Fabric based electrodes were glued to micro-slide glass plates (for mechanical support), and Cu2S was deposited on a brass foil. All the remaining CE materials were deposited on FTO substrates.

19

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 22

TABLES Table 1 Raman data of carbon nanostructures.

-1

G-band

ID/IG

CE

D-band (cm )

Soot

1347

1584

1.123

SWCNTs

1345

1583

1.654

F-MWCNTs

1350

1588

2.109

C-Fabric

1346

1598

2.282

-1

(cm )

Table 2 Solar cell parameters of cells with 0.1 M Na2S electrolyte, with TiO2/CdS as the photoanode, exposed cell area: 0.08 to 0.10 cm2, under 1 sun illumination (100 mW cm-2) with the listed counter electrodes.

Counter Electrode

Voc

Jsc -2

FF

PCE

Average η

(η, %)

(%)

(mA cm )

(V)

SWCNTs

6.59

0.791

38.75

2.021

1.92

WO3

6.81

1.077

28.63

2.1

2.0

PEDOT

8.892

0.759

41.40

2.79

2.76

Cu2S

12.56

0.723

31.70

2.88

2.757

Soot

8.842

0.965

34.50

2.954

2.93

WO3-x

7.9

1.04

44.54

3.66

3.338

F-MWCNTs

9.99

0.852

44.38

3.78

3.618

C-Fabric

10.04

0.924

42.78

3.97

3.893

C-Fabric/ WO3-x

8.86

0.951

51.51

4.6

4.45

20

ACS Paragon Plus Environment

Page 21 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Table 3 Redox potentials (from LSV plots) and charge transfer resistances (from EIS plots) of counter electrode films.

Counter Electrode

E (V vs. NHE)

ipeak(red)

RCT (Ω) -2

(mA cm ) SWCNTs

-0.617

–0.676

247

WO3

-0.612

-0.884

202

PEDOT

-0.611

-0.93

160

Cu2S

-0.603

-0.5809

152

Soot

-0.600

-0.866

125

WO3-x

-0.586

-0.8836

110

F-MWCNTs

-0.576

-1.4

45

C-Fabric

-0.538

-1.32

38

C-Fabric/WO3-x

-0.558

-1.01

7

21

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 22

Graphical Abstract

22

ACS Paragon Plus Environment